I have been going through my older observations and making them research grade when possible using the Data Quality Assessment. Realised though that quite a few of the invertebrates are at family level and won’t go further, which seems common in quite a few invert families for field observation.
This limits me to making an observation casual or leaving it open as needs ID.
e.g.
Cecidomyiidae
https://www.inaturalist.org/observations/27885329
It seems strange to me that family level identifications can’t reach research grade.
(This has been mentioned elsewhere, but there wasn’t a response to the family-level issue
https://forum.inaturalist.org/t/family-and-genus-level-ids/4083 )
Particularly as this impacts the computer vision trawling if I understand correctly - is that right?
Its only 100+ RG obs of genus level or higher which will impact CV training?
The CV is already limited enough on many inverts, but if its not able to suggest inverts which are at family level, then this is kind of tying its hands behind its back so to speak…so I would just like to ask what the reasoning is behind this?
Also of note in this instance, most gall midge images are of the host plant. So in this instance, it seems like there will be a heavy weighting to species level ID using plant hosts and no input of the diptera images as they are at family level…would this make it even more unlikely that field observations of the diptera themselves will be autosuggested as cecidomyiidae?